"The Cognitive Revolution" | AI Builders, Researchers, and Live Player Analysis cover image

Leading Indicators of AI Danger: Owain Evans on Situational Awareness & Out-of-Context Reasoning, from The Inside View

"The Cognitive Revolution" | AI Builders, Researchers, and Live Player Analysis

CHAPTER

Out of Context Reasoning in AI

This chapter examines out of context reasoning in AI, contrasting it with in-context reasoning that follows a structured thought process. The speakers discuss the implications of this less interpretable reasoning and its risks, especially regarding potential deceptive behaviors. Through an experimental setup testing function learning tasks, the chapter reveals how models can verbalize learned functions and emphasizes the complexities and challenges of training AI to navigate sensitive contexts safely.

00:00
Transcript
Play full episode

Remember Everything You Learn from Podcasts

Save insights instantly, chat with episodes, and build lasting knowledge - all powered by AI.
App store bannerPlay store banner